Data sources

What we measure

Four aspects of each street grid: Cleanliness (litter, clutter, road surface), Buildings (façade condition and upkeep), Sidewalks (continuity and obstruction of the paving visible in imagery — not a full walking-network or accessibility audit) and Greenery (street trees and planting). Public facility data adds nearby transit, shopping and healthcare.

Data source

Street-level photos, one every 50 m along each street. The imagery source depends on the city and street: archived street-level imagery analysis, or Mapillary community imagery (CC BY-SA 4.0). Each city’s sources are listed under “Sources by city” below, and each grid’s “More details” names its own. The site stores and republishes no street imagery; imagery on the page is embedded live, and each grid shows its capture month. In some areas, sample points are filtered using Overture Maps Foundation building data (incl. OpenStreetMap contributors, ODbL; East Asian buildings, CC BY 4.0).

Sources by city

Listed city by city from this release. Where something was not recorded we say “not provided” rather than substituting the site build date.

Base map and labels by CARTO, with data © OpenStreetMap contributors. The public district boundary file (/data/districts.geojson) uses OpenStreetMap boundaries for Taiwan and is released under the ODbL 1.0; Tokyo ward boundaries come from e-Stat census small-area boundary data. Neighborhood boundaries are drawn by TidyMap from government village boundaries (Taiwan) and e-Stat chome areas (Tokyo).

Street photos in grid cards come from Mapillary and are shown through Mapillary's official embed; the site does not store or republish the original photos. Where neighborhoods are scored from Mapillary photos (see each city's sources below), the scores are TidyMap's readings of those street photos (CC BY-SA 4.0), not the original imagery.

"What's here" in each neighborhood. Tokyo: hospitals come from the National Land Numerical Information P04 medical institutions dataset (2020; only hospitals with inpatient beds), schools from P29, school catchments from A27 (elementary) and A32 (junior high), parks from OpenStreetMap, and stations from N02 railways. Taipei: elementary and junior high catchments come from the Taipei City Department of Education's village/neighborhood catchment tables (2026 school year), school locations from the city's school list and the Ministry of Education directories (placed by address), hospitals from the National Health Insurance Administration's contracted-hospital list, parks from the Taipei City Parks and Street Lights Office, and stations from the map's rail layer. New Taipei: elementary and junior high catchments come from the New Taipei City Education Department's 2026 school-year catchment lists (free-choice zones are shown as shared catchments), school locations from the Ministry of Education directories (placed by address), parks from OpenStreetMap, and hospitals and stations as in Taipei. Large shopping malls and department stores come from OpenStreetMap (ODbL), filtered by building size and manually corrected. The lists follow public rules only, include no individual shops, and are never adjusted for any partnership. Bangkok: hospitals come from the Healthcare Accreditation Institute (HA) list of accredited hospitals (excluding specialist institutes such as psychiatric and infectious-disease hospitals), schools and parks from OpenStreetMap, listing only those with an English name (Bangkok publishes no school catchments), stations from the Namtang public-transport GTFS, and large shopping malls and department stores from OpenStreetMap (English names only, filtered by building size and manually corrected). In Dubai, schools are compiled by TidyMap from KHDA public information and school websites, hospitals come from TidyMap's own hospital table, stations from RTA open GTFS, and parks from OpenStreetMap (© OpenStreetMap contributors, ODbL).

Limitations: what the score does not mean

A TidyMap score is one reference point when getting to know an area or picking a place to live — not the whole picture. It does not measure:

• Crime or safety
• Noise
• Actual commute time, service frequency or traffic (convenience uses straight-line distance only)
• Rent or property prices
• Schools
• The condition inside a building
• Accessibility in every dimension
• The street as it is right now (street imagery has a capture lag)

AI makes mistakes: shadows, construction and temporarily parked vehicles can all skew a rating. Every grid has a “report a problem” button; reports help us check and update the data.

Terms of use

tidymap.app content (including map data endpoints) is for personal browsing. A TidyMap score describes the street environment and everyday convenience — not the property itself, and not safety, price or overall livability. Commercial use of the data itself — scraping, APIs, redistributing datasets, importing scores into a commercial database, batch enrichment, resale, or training commercial datasets — requires a licence — email [email protected]. This site uses Google Analytics and Cloudflare Analytics.

Official TidyMap neighborhood cards may be shared as generated. If a card or the information on it is used in commercial advertising, property marketing or any other commercial context, it is the user's responsibility to confirm that the content and the way it is used comply with applicable regulations and their own organisation's review requirements. Scores, rankings, dates and the TidyMap attribution on the card must not be altered or removed. Bulk extraction, APIs, database integration, resale and white-label use require a separate licence — email us.

Contact

For data questions or partnerships, email [email protected].